A bipartisan group of US lawmakers moved forward this week with legislation designed to prevent Chinese-made humanoid and quadruped robots from gaining a foothold in America's critical infrastructure, warning that Beijing is deploying the same industrial strategy it used to dominate the global drone market — and that connected robots could be exploited for espionage and physical sabotage.

What the GUARD Act Proposes

The legislation, known as the GUARD Act, was the centrepiece of a House Energy and Commerce Communications and Technology Subcommittee hearing held on 25 July. The session examined a range of bills covering communications security, artificial intelligence, broadband deployment, and illegal robocalls. Lawmakers from both parties argued the United States must act before Chinese robotics achieve widespread adoption across American industry.

Key Warnings From the Hearing

Committee Chairman Brett Guthrie noted that Congress had previously acted against telecommunications equipment and drones from Chinese manufacturers. He cited 'recent reports of a China-linked plot to cripple communications networks' as evidence that threats to American infrastructure remain active, adding that lawmakers are 'committed to protecting our critical infrastructure and safeguarding Americans' data.'

Michael Robbins, President and Chief Executive Officer of the Association for Uncrewed Vehicle Systems International (AUVSI), urged swift action. 'The same PRC playbook used for drones is now aimed at robotics,' Robbins told the panel. 'The humanoid and quadruped market is still relatively new.' He warned that Chinese firms already account for 'roughly 90 percent of global humanoid shipments' and that American research institutions have shifted toward 'subsidised, artificially low-cost Chinese systems.'

Robbins drew a stark distinction between conventional cyber threats and robot-specific risks. 'A compromised laptop can expose data. A compromised robot can expose data and also move, surveil, and act physically inside a factory, a hospital, a power plant, or a military base,' he said. 'Every such system is therefore a Trojan horse on American soil.' He also testified that connected systems are 'known to be directing data directly back to the PRC' and that robots are 'mapping, surveilling, and collecting audio through sensors,' with software updates potentially enabling physical disruption.

The AI Dimension

Former White House technology adviser Lindsay Gorman cautioned that artificial intelligence has fundamentally altered the cybersecurity landscape. 'AI is massively expanding the cyberattack surface and risks tipping the offense-defense balance in favour of the attacker,' she said. Gorman argued that the US needs 'lower-cost alternatives for export and a vibrant open-source AI ecosystem at home' to compete with inexpensive Chinese AI models being promoted globally, warning that many countries could adopt them purely on cost grounds.

The Case for Allied Coordination

Robbins also pressed for closer cooperation with trusted partners, identifying Japan, South Korea, Taiwan, Israel, Australia, and European allies as key collaborators. 'We are much stronger when working together to take on China's authoritarian scale than when any single democratic nation tries to take on China's authoritarian scale alone,' he said. He cautioned that delaying policy action would deepen dependence on Chinese technology and make future restrictions considerably more costly.

Broader Agenda at the Hearing

Beyond robotics, the subcommittee also examined proposals to deploy AI to strengthen communications networks, improve federal broadband coordination, and crack down on illegal foreign robocalls — threats that witnesses increasingly linked to organised criminal networks operating overseas. The hearing signals a broadening legislative push to address technology security risks across multiple sectors simultaneously.